A review of the application research on artificial intelligence-assisted surface acoustic wave sensors in biomedical engineering

Surface acoustic wave (SAW) sensors offer high-frequency piezoelectric transduction, rapid response, miniaturization, and potential passive or wireless operation, making them attractive for biomedical sensing. However, their practical performance is affected by complex biological matrices, environmental drift, cross-sensitivity, device variation, and weak or nonstationary signals. This review links the physical sensing mechanisms and intrinsic performance of SAW devices with the incremental contributions of artificial intelligence (AI). SAW performance is summarized for disease-biomarker detection, physiological monitoring, and human-motion sensing, while AI-assisted approaches are evaluated in terms of device optimization, drift and interference compensation, multidimensional feature extraction, and sensing-computing integration. AI can improve information yield and system robustness without replacing the underlying piezoelectric transduction or sensing interface. Key challenges include limited datasets, calibration transfer, model generalization, biofouling, matrix shifts, computational constraints, and insufficient independent or clinical validation. Future development will rely on physics-informed design, digital twins, robust biointerfaces, adaptive multimodal compensation, edge intelligence, and standardized validation.